If you want to have this newsletter mailed to you or you want to make comments/suggestions about the format/content then send an email to acedb@sanger.ac.uk.
The Sanger Centre has had to tighten up security on its ftp and web sites. This has meant that some users have had trouble downloading via ftp because their own ftp is not configured to use passive mode which is the only supported mode at the Sanger Centre now.
Do please remember that you can download everything from our web page for downloads using either ftp or http protocols, the latter will avoid any problems you may have with ftp.
The server can now be run in the foreground in "silent" mode, i.e. no messages are sent to the terminal. Specify the "-silent" option when starting the server. This option was requested to make it possible to run the server for extended periods of time in an xterm without huge numbers of messages being output to the terminal.
There has been some confusion over the valid range of port numbers for the socket server. IANA (Internet Assigned Numbers Authority) has set aside certain ranges of port numbers for certain uses with sockets:
Port No. range Usage
-------------- -----
0 - 1023 "well-known" ports, you need super user authority
to use these and they are reserved for ftp, http,
X windows and basic operating system services.
DO NOT USE THESE.
1024 - 49151 "registered" ports, if you setting up the server to
be run by inetd you should use one of these.
But you will need to check in /etc/services
that your number does not clash with one already there.
49152 - 65535 "dynamic" ports, are for dynamically assigned/ephemeral
use. You should use one of these if you are just trying
out the server from an xterm in the foreground.
You should not choose a port larger than 65535. If you do, the port number may be truncated in a way that leaves it in the range 0 - 1023 which are reserved for superuser use for well-known services. Numbers larger than 65535 are generally reserved for RPC and not available for sockets.
In the past when acedb crashed it produced a core dump. Unfortunately this dump is often useless because acedb attempts to clean up the database before the core dump by setting up a special routine called a signal catcher. This means that the core dump records what was going on when acedb was cleaning up, not what caused the crash.
Now, by default, no core dump will be produced. Instead, if acedb crashes in a way that you can reproduce, you should restart it with the "-nosigcatch" command line option, e.g.:
tace -nosigcatch your_database
This will tell acedb not to try to clear up when it crashes and the operating system will then produce a useful core dump when the problem reoccurs. You should keep this core dump and let the acedb developers know about it.
If you are using xace, you can also turn signal catching on and off from the Admin menu on the main window.
The grey ramp printing code was preforming a function call per pixel, this has now been fixed which should speed things up a lot.
When the aceOut package was introduced it unfortunately encouraged some erroneous use of the aceOutPrint() call (which should really be called aceOutPrintf()). The problem occurs when programmers do this:
char *some_string = "blah, blah, resulting in a large % gain, blah, blah" ;
aceOutPrint(some_string) ;
char *some_string = "blah, blah, resulting in a large % gain, blah, blah" ;
aceOutPrint("%s", some_string) ;
All occurences of this in acedb code have now been fixed and the header for the aceOut package updated to warn programmers about this problem. A longer term fix will be to introduce an aceOutPrintf() function to take the place of the aceOutPrint() call, which will become a routine to print simple strings.
Dotter was crashing periodically, this was because of a problem passing the data over to the forked off dotter via a pipe. This should now be fixed.
Jean Thierry-Mieg discovered a serious bug which affected the use of aliases with acedbs cache of the tags present in an object. The cache is used to speed up queries since it makes it possible to see if an object contains a particular tag without actually having to load the object.
The code was not de-aliasing names and so would report a tag missing from an object because it did not look up the true object underlying the alias in the cache. This should make aliases much more reliable to use.
The library that contains the GD package for producing acedb images used to be called libgif.a, it has now been renamed to libgd.a because of a name clash with a library on the new Mac OS-X system. This more accurately reflects the library's contents in any case since it is based on the GD package.
Due to problems with our build system and some large scale and unfinished changes to introduce Smap (see January issue) we were not able to produce a December monthly build. Sorry !